Data is logged from a tester to represent events occurring on the tester. For example, there is an event for starting to test a lot. There is an event for starting to test a device. There is an event for execution of an individual test. The order of the events is important and provides context information. When an event is logged that signals the beginning of testing for a lot, all device test results that are logged after that point, and prior to an event signaling the end of a lot, are considered part of that lot.
Due to complexities of testing systems, and to allow flexibility in using the system, it is possible to have “missing events”. These missing events in turn cause missing context information for the testing system. These missing events must be addressed for each of the processes prior to formatting test data. For example, to format data to a text report and to a binary STDF file, both of these processes need to compensate for the missing data.
In the past, each individual data formatter would provide its own solution for the missing data. For example, unique data relationships may be provided from one data formatter to another. Output data from some or all of the data formatters could be provided with holes or missing relationships. One or more of the data formatters could potentially crash.